RIGHT-VENTRICULAR END-DIASTOLIC VOLUME INDEX AS A PREDICTOR OF PRELOAD STATUS IN PATIENTS ON POSITIVE END-EXPIRATORY PRESSURE

Objective: To evaluate the clinical utility of right ventricular end-d iastolic volume index (RVEDVI) and pulmonary artery occlusion pressure (PAOP) as measures of preload status in patients with acute respirato ry failure receiving treatment with positive end-expiratory pressure. Design: Prospective, cohort study. Setting: Surgical intensive care un it in a Level I trauma center/university hospital. Patients: Sixty-fou r critically ill surgical patients with acute respiratory failure. Int erventions: All patients were treated for acute respiratory failure wi th titrated levels of positive end-expiratory pressure (PEEP) with the goal of increasing arterial oxygen saturation to greater than or equa l to 0.92, reducing FlO(2) to <0.5, and reducing intrapulmonary shunt to less than or equal to 0.2. Serial determinations of RVEDVI, PAOP, a nd cardiac index (CI) were recorded. Measurements and Main Results: Tw o hundred-fifty sets of hemodynamic variables were measured in 64 pati ents. The level of PEEP ranged from 5 to 50 cm H2O (mean 12 +/- 9 [SD] cm H2O). At all levels of PEEP, CI correlated significantly better wi th RVEDVI than with PAOP. At levels of PEEP greater than or equal to 1 5 cm H2O, CI was inversely correlated with PAOP, but remained positive ly correlated with RVEDVI. Conclusions: CI correlates significantly be tter with RVEDVI than PAOP at all levels of PEEP up to 50 cm H2O. RVED VI is a more reliable predictor of volume depletion and preload recrui table increases in CI, especially in patients receiving higher levels of PEEP where PAOP is difficult to interpret.